January 2012 - users-list

[MonetDB-users] [PATCH]: python-monetdb does not support RIPEMD160
by W-Mark Kubacki 09 Jan '12

09 Jan '12

Hello! Although MonetDB supports RIPEMD160, python-monetdb (version 11.5.9) does not. Patch [1] fixes that. Unfortunately Python's hashlib.algorithms and dir(hashlib) is "incomplete" in this regards and lists only algorithms which are guaranteed to be supported. Yet more are actually implemented. The patch moves the responsibility for checking whether RIPEMD160 (and others) is supported to hashlib, retaining the NotSupportedError. -- Mark Kubacki [1] https://raw.github.com/wmark/ossdl-overlay/master/dev-python/python-monetdb…

2 1

[MonetDB-users] Max Connection Limit
by Joseph Brower 06 Jan '12

06 Jan '12

Right now we are hitting the limit of 64 connections to MonetDB. We're working on some middleware that will consolidate those into only 2 or 3 connections, but until we finish it, we need to increase the limit. How do we go about doing that? Thanks, Joseph Brower

3 7

[MonetDB-users] Error in SELECT from a function via Mapi
by Iain Buchanan 02 Jan '12

02 Jan '12

Hi, I'm having problems calling a function via the Mapi interface (using Visual C++ under Windows and the 15th November download). It works fine when I call it directly, but fails when I call via Mapi. I create the function: create function tick.dummy(a integer) returns table (entry integer) begin return (select 1); end; I can then call this directly from mclient without any problems: sql>select entry from tick.dummy(1); +-------+ | entry | +=======+ | 1 | +-------+ 1 tuple (1.074ms) sql> However, if I call via Mapi I receive an error: SELECT: no such unary operator 'dummy(tinyint)' I'm using the following sequence of API calls: MapiHdl handle = mapi_prepare(connection, "SELECT entry FROM tick.dummy(?)"); int param(0); mapi_param_type(handle, 0, MAPI_INT, MAPI_AUTO, &param); if (mapi_execute(handle) != MOK) { // Error here } The above seems to work fine for a general SELECT statement, but fails for one with a function. Is there a different syntax for calling via Mapi? Iain

1 0

[MonetDB-users] ENUM style table with JOIN versus automatically deduplicated VARCHAR
by Stefan de Konink 02 Jan '12

02 Jan '12

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hi, MonetDB 'deduplicates' strings, is there any benchmark that outlines the performance difference of an ENUM with a low number of (string)items versus the use of a VARCHAR field? The technical question would be; Is it more efficient to do: select * from my_table, my_enum where mytable.my_enum_id = my_enum.id and my_enum.name = 'Something'; or select * from my_table where my_enum_varchar = 'something'; I can imagine that the following would out perform a string comparison, but implies hardcoded values. select * from my_table, my_enum where mytable.my_enum_id = hardcoded; Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.18 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEAREKAAYFAk78fnkACgkQYH1+F2Rqwn3BUQCeOUtPIGzcHT1xpnsF4n78BQ/1 JWAAnjnzPeT8D7CtsjYgAiGg/WRIgNbQ =989f -----END PGP SIGNATURE-----

2 1